Retell Sales

In yesterday’s note (click here to review) we examined the Census Bureau’s data on Retail Sales.  In short, they were weak.  The headline tally declined 1.2 percent (the largest drop in nine years), and minus signs accompanied most of the components. 

This iteration left Atlas puzzled for a couple of reasons.  First, headlines throughout the holiday shopping season were positive for brick-and-mortar stores and online merchants alike.  Secondly, and perhaps more importantly, the trend for this indicator had been moving steadily higher, so the sudden contradiction was quite surprising. 

Fortunately, retail sales are released frequently and subject to revisions.  We’ll get another look at this data set soon with January’s release next month.  The next couple of months should help clarify whether December’s figures are simply an outlier or something more troubling.  If it is indeed a one-off, it could be an issue with seasonal adjustment.  These are a statistical tool designed to remove patterns associated with the time of year.  They are often useful adaptations which allow easier comparisons between different seasons (e.g., comparing holiday season to summer months).  However, sometimes the adjustment is overly strong and can cause figures to move incongruently to the trend.

In case you are wondering, Atlas isn’t willing to throw in the towel on this indicator.  The year-over-year trend is still positive, up 2.3 percent since December 2017.  Seasonal adjustments are a tricky beast, and we are not too confident in this iteration, feeling it should have been stronger than this report shows.  We’ll watch the rate of change in the trend.  If we see something telling, we’ll tell and possibly retell you about it here.